Discussions of AI and robotics often focus on the question of whether AI can surpass human intelligence, and on some rather extreme visions of the future. But the more mundane implications are already in play, and they are actually not so mundane after all: we have a scientific revolution under way.
Scientific revolutions are rare turning points in the progress of human understanding. Wouldn’t it be exciting to witness a scientific revolution up close? Wouldn’t it be sad to live through a scientific revolution without noticing? But that is exactly what is happening. We are living through a scientific revolution, and nobody has noticed. Robotics and artificial intelligence (AI) are changing how we see ourselves, and how we see intelligence in general. Unfortunately, the current hoopla is oblivious to the science, focused instead on the technology — chatbots, deep fakes, humanoids, and so on. Even those attuned to the science have not noticed that it rises to the level of a scientific revolution.
What is a scientific revolution? We humans tend to think that we are at the center of things — the center of the universe, or the pinnacle of the animal kingdom, for example. But then, every few hundred years, we get a reality check — a bucket of freezing cold water in the face. Humans are not the center of everything, after all. Science historians refer to these reality checks as de-centering events. There are other ways to define scientific revolutions, notably the paradigm shifts popularized by Thomas Kuhn, but I am adopting the higher standard, defining scientific revolutions as de-centering events. Under this definition, only two previous scientific revolutions are widely recognized. Let’s review …
Revolution #1: the Copernican Revolution. Humans used to occupy the center of the physical universe. Earth was at the center, and everything else circled around it — the sun, the moon, the planets, and the stars, all embedded in celestial spheres. But then Nicolaus Copernicus came along, about 500 years ago, and touched off the Copernican Revolution. We are not the center of the universe, after all. The earth orbits the sun. And things just get worse from there. The sun isn’t the center either. Our solar system is out on the rim of a galaxy, one of many galaxies. There is no center! This was a dramatic and fundamental change in the accepted world view, and it took us about a hundred years to adjust and accept this radical new idea. But we did adjust and accept, and we grew up a little.
Then, a few hundred years later, it happened again! Another de-centering event, another bucket of cold water in the face. Revolution #2: the Darwinian Revolution. Humans used to be the center of the biological universe. No, that is an understatement, we even stood apart from the biological universe. Humans were conscious and intelligent; animals were not. Then Charles Darwin came along, about 150 years ago, and touched off the Darwinian Revolution. Humans are apes, descended from ancestors we hold in common with chimps and other apes, incrementally coming to our present form through the processes of natural selection. This may be premature, but let’s say that we adjusted, we survived, we learned, and we grew up a little more.
And now, it’s happening again. Another dramatic de-centering event. Revolution #3! We humans, along with our animal cousins, seemed to be the only intelligent agents in the universe, but that has ended. Robotics and artificial intelligence (AI) are creating new intelligent machines.
By designing our own intelligent machines, we can explore the nature of intelligence in ways heretofore unimagined. There is no limit to the number and variety of intelligent agents, at least none we can discern now. What we previously believed to be the entirety of intelligent agents will instead occupy just one corner of this new universe.
Among many of us, the new view has already taken root. Many in philosophy, cognitive science, AI, robotics, and other fields believe that the brain is a type of computer. If true, the sharp distinction between intelligence and non-intelligence begins to look not so sharp. As an example, John McCarthy (one of the founders of AI) famously declared that even a thermostat has beliefs and goals. Specifically, it can have any of three beliefs: “The room is too hot,” “the room is too cold,” or “the room is OK.” The thermostat’s goal? “The room should be OK.” (McCarthy 1979) There is, of course, a vast difference between a thermostat and the human brain. But perhaps that gulf could be spanned by a range of artificial intelligences, at least in principle and maybe someday in practice, just as evolution spanned the gulf from single-celled organisms to humans, defining a gradual progression from low to high intelligence..
Here is the question that puzzles me: Why has nobody noticed that we are living through a scientific revolution? Everything I say above is familiar to and accepted by many working in the related fields. And yet, I can find almost no references to it as a scientific revolution. Why? Perhaps it is a question of time. Existing scientific mindsets are not replaced easily or quickly. People are invested in the ideas of their early training, and in ideas developed over a lifetime of work. It took decades, or even centuries, before Copernicus’s theory was fully accepted. Even today there are holdouts — the flat-earthers for example. In Darwin’s case, the evolution of species was generally accepted in his lifetime, but natural selection was not accepted until much later, and there are still quite a few holdouts.
So, many will duck this bucket of freezing water, stick with the old ideas about intelligence, and miss the fun. But I have to admit: the flat-earthers seem to be having fun, too. Perhaps the resistance to revolution #3 will develop into an equally fun movement, with demonstrators waving signs: Thermostats don’t think!
What will future science historians call this revolution? Perhaps, with the historical perspective, they will look at it more broadly, and associate it with computation in general, rather than just AI and robotics. It might be called the Computational Revolution, or the Turing Revolution.
One more question: Will this be the last scientific revolution, or are more de-centering events possible? Note that the scientific revolutions so far have become progressively more intimate. Copernicus addressed the ground at our feet and the heavenly bodies. Darwin was a step more intimate, addressing our bodies and our ancestry. Revolution #3 is taking another step, addressing our most intimate selves: our minds. Might we find that there is yet a fourth step required to understand the mind in full? Or will Revolution #3 be humankind’s final step to achieve proper humility?
Reference
McCarthy, J. Ascribing mental properties to machines. Stanford AI Lab Memo AIM-326. March 1979.
I’ve noticed that we are living through a scientific revolution myself, though when I told people, I framed it along the lines of “Similar to how Newton discovered gravity, we are in the middle of discovering how consciousness actually works.” I suppose I’ve already taken the breakthrough in intelligence for granted!
I believe the reason why people do not notice we are in the middle of a scientific revolution is because we are witnessing the revolution in relatively slow motion (in years). The intelligence is strongly tied to the amount of available compute measured in orders of magnitude. The scale of compute required and the rate of growth of compute means that to the layperson they only see the step changes in capabilities every few years, when some large company successfully trains the next largest model.
I like to think of all this in terms of # of human brains worth of compute per dollar because it gives me a way to project when we will reach X capability. To help anchor ourselves, Deepseek R1 was on the order of 500B parameters, which is not too far from a mouse brain of ~100B parameters! IIRC, human brains might have 100T parameters. Some interesting milestones then are what happens when you can hire 1 smart person for $1 an hour, or 10 smart people e.g., a small startup, or 10,000 smart people e.g., a global corporation or an entire research community.
Great minds think alike!
And I am not talking about by own mind but about Bruce Mazlish’s. In his 1967 article “The Fourth Discontinuity” he made similar arguments about Copernicus, Darwin — then he adds Freud as another de-centering revolution — and the AI revolution.
https://www.jstor.org/stable/3101522?seq=1
He turned the article into a book later:
https://yalebooks.yale.edu/book/9780300065121/the-fourth-discontinuity/
But I don’t think that the AI revolution is here (yet)…
Best,
Oliver
Thanks for posting this. I wasn’t aware of Mazlish’s paper or book. He doesn’t use the phrases “scientific revolution” or “de-centering event” but uses the terms “discontinuity” and “ego-smashing” referring specifically to Copernicus, Darwin, Freud, and then AI, so there is no doubt. I have to read it more carefully, though. He actually doesn’t refer to AI, by name. And he wrote this in 1967!
Great post! We have one in the pipe on our blog about why Robotics is a pre-paradigm field, based on Thomas Kuhn’s the structure of scientific revolutions, stay tuned! I think POMPDs and handling partial observability/active information gathering is an important piece missing from the discourse right now. But all of that is at a lower level than what you are talking about here – we are living in interesting times, and it’s so exciting to be a part of the community right now!
I enjoyed this framing. The de-centering lens is clarifying, and the progression from Copernicus to Darwin to AI does capture something real about how these revolutions work. But I want to push on the “humility” angle a bit, because I think there is a tension worth exploring.
Each revolution you describe didn’t actually make us humble. It made us relocate our arrogance. After Copernicus, we shifted from “we’re at the center of the cosmos” to “but our reason can comprehend the cosmos.” After Darwin, we pivoted from biological uniqueness to consciousness and culture as what makes us special. Now AI is threatening those, and we’re already hunting for the next redoubt. (I’ve seen, and wrote about, “embodied experience” and “authentic emotion” floated as candidates.) Real collective humility might require accepting that there’s no final “special” thing, that intelligence and meaning exist on continua we don’t occupy the endpoints of. As Stefanie mentioned, we are thinking about this from a different framing as well.
Your observation about public fixation on chatbots and humanoids reminds me of Plato’s cave. The shadows on the wall are more legible than the light source. People find it easier to react to AI hype than to sit with the harder question you’re raising: what does it mean for intelligence to exist on a continuum? The technological discourse is a shadow puppet show, and the light source is the scientific revolution happening live around us.
I’ll be curious to see what historians call this revolution in 100 years. “Turing Revolution” has a nice ring to it. Though, if we’re being honest, they might just call it “the thing humans almost missed because they were too busy arguing about chatbots.”
Nice comments
Two thoughts:
1 A key distinction between the first two technological revolutions and the current AI-driven revolution is the rate and breadth of information diffusion. Earlier revolutions were initially limited to small expert communities and propagated gradually through society. In contrast, contemporary AI technologies are widely visible and accessible from their inception, enabling near-instantaneous global awareness and adoption. This accelerated diffusion raises important questions about the tempo and magnitude of societal transformation.
2 These tools clearly make us more productive and often improve the apparent quality of our work, whether in writing or problem solving. But they also offer us the opportunity to outsource thinking itself. The risk is not that we become less capable overnight, but that, over time, we quietly lose the habit of pushing through hard ideas—the very process that builds judgment, intuition, and real expertise. I wonder: did engineers once believe that not learning how to draft would deuterate their skills? I certainly used to memorize all of my phone numbers until I got a cell phone. I think I am better off with the cell phone.
Scientific Revolution #4: de-centering of self, all minds become one mind 🧘😌🙏🏻
.
Ancient cultures have always assumed the existence of nonhuman (often incorporeal) intelligences: demons, genies, fairies, and the like. These intelligences varied from aligned to non-aligned to indifferent. I wonder if the revolution you describe returns us to our “natural” worldview and what the implications of that might be.